A sense of the term anonymous with which I had previously been unfamiliar


I was filling out one of those online satisfaction surveys, and in the introduction, it reassured me that

This survey is anonymous.

The first question on the survey:

Enter the support request (SR) number below.

Yeah, because they'll never be able to trace the support request number back to me.

(I suspect that the reason for this contradiction is that the organization that wanted to conduct the survey used a site that supports either anonymous surveys or tracked surveys, and they opted for the anonymous survey for whatever reason—maybe it's cheaper?—but they actually wanted a tracked survey, so they made the tracking number part of the survey itself.)

Comments (18)
  1. Alexander Grigoriev says:

    The number just makes sure the survey is only filled once. And to protect it against flash mob pranks.

  2. Gavin Greig says:

    In a previous job, a colleague and I completed an "anonymous" survey about some training given by employees of another authority regarding the software they wrote and which we used.

    Not long later our manager, who hadn’t been there, asked us why we’d given negative responses to the survey. Clear anonymity fail. Actually, I don’t think they were terribly negative – but apparently everyone else had praised them to the skies where we’d tried to be objective (see Raymond’s post yesterday: http://blogs.msdn.com/oldnewthing/archive/2010/02/10/9960892.aspx). There were no negative consequences for us – unless you count world-weary cynicism ;-)

  3. rw says:

    @a random passerby:

    Your method is equally trustworthy than. Alexander’s. Who says they store the support request number together with the survey data?

    If they want relate the data, they can just store the "randomly-generated-but-unique number – support request number – relation".

    You just cannot be sure.

  4. John says:

    Obviously the survey was done this way in order to serve you better.

  5. Market Research IT Analyst says:

    "The real way to do that is to send an invite with a randomly-generated-but-unique number. Internally, the number is never associated with the person filling out the survey"

    I can assure you that the number is associated to a person at some point, even if that point is only when the email is generated to for person with that number. Most likely, if the survey invite is sent to you, they will have the ability to know what you said.

    There is too much to be said on this topic for this space. Suffice it to say that your answers are only as anonymous as the entity conducting the survey is trustworthy – or competent.

  6. Miral says:

    Clearly, you should have typed "the support request (SR) number" (literally) into the box.  That’s the only way of reconciling the two. :P

  7. Some Guy says:

    They say: this survey is anonymous

    They mean: you expect us to use your contact details against you, and we’re so retarded we still need to do a survey to find out whether or not you trust us

    As a rule, a survey is either being done by real scientists, or by a manager who doesn’t know what he’s doing but wants to look busy. A customer survey is another way of saying that management is hiding on the 87th floor two states away, and doesn’t even want to listen to the opinions of people who actually deal with customers every day.

    In reality, people who do the work already know the difference between "I wish I wasn’t here, but there’s no competing store that’s any better" and "I’m happy to come here and spend money" by the time the customer has walked in the door. If your beauracracy insulates the decision makers from this information, no amount of surveys will actually help. And if you have a large enough company, you can quickly and easily teach your employees that actually caring about problems is outside their job responsibilities, and replace any who do show initiative.

    Fortunately for management, by the time you get to be that big, you can have all the customer focussed group initiatives you want, and do pratically nothing to shift your customers away from the "eh, whatever, its just another BigCo branch" attitude that an accurate survey would have discovered, plus they get to tell their superiors how customer focussed they are without having to meet the customers.

  8. Many by-default-anonymous data collection mechanisms allow you to volunteer your identity (like if you want someone to be able to contact you for clarification on your feeback)

  9. a random passerby says:

    @Alexander:

    I hope that was sarcasm or a joke. The real way to do that is to send an invite with a randomly-generated-but-unique number. Internally, the number is never associated with the person filling out the survey, but is only used to make sure the survey can only filled out once. And yes, if a person gets two invites for some reason, they get to fill out the survey twice.

    A "support request" number that *is* associated with a name completely breaks the supposed anonymity of the survey.

  10. Scott says:

    This is why I always sign "anonymous" surveys and complaints. I may as well make it easy for them to trace it back.

  11. Esko says:

    I’ve seen this several times, the greatest example I can think of is ESPAD (the European School Survey Project on Alcohol and Other Drugs) and yes, naming that is really important in this context. The questionnaire asks questions such as which drugs you have taken, who do you (as a minor) buy your alcohol from, how many friends you have, what are their full names and which ones of them drink alcohol/smoke etc.

    So, when I was still in school I had to participate in filling the ~120 question questionnaire yearly for four years. Mind you that the very first paragraph in the introductory text says that it’s not mandatory yet we weren’t allowed to return blank forms. Oh well, at least it’s anonymous just as the same info blurb says, right?

    For context, the questionnaires are handed out in envelopes with your name on them and which contain the actual questionnaire and another "blank" envelope for returning the questionnaire. The name is also coupled with a number.

    On the fourth year we weren’t stupid anymore so instead of answering this "anonymous" questionnaire we decided to dissect the pile of papers in that envelope and after a while decided that the promise for anonymity is a complete fluke:

    • The questionnaire had an unique barcode
    • The return envelope had an unique number code inside it at the bottom corner which matched with the one in the envelope the questionnaire came in

    • The first two dozen questions were enough to distinguish you from a group ("What is your school?", "What is your class?", "What is your gender?", "How tall are you?", "How much do you weight?" etc.)

    • The very last page of questionnaire contained your name.

    We gathered a small report about these issues, represented them to our teachers and after some bureaucracy didn’t have to fill a single one of those questionnaires ever again.

  12. Paul says:

    I had a postal anonymous survey sent.

    On the back page of a letter with my name and address on.

    I didn’t fill it in.

    I then got a letter telling me I did not complete the survey and needed to.

    Nice anonymous one!

  13. Words Have Meanings says:

    @Dont matter

    Retard – General definition:

    To slow or hinder.

    A false statement that says X and does !X would definitely qualify as retarded.  It takes time to be identified as a misleading statement and prevents other activity while that discovery is being made.  Sounds pretty much spot on with the definition of retard.

  14. > I then got a letter telling me I did not complete the survey and needed to.

    Maybe *nobody* completed the survey so they sent everyone the nag letter.

    Or maybe everybody who did complete the survey volunteered indentifying information.

  15. Johan says:

    I worked for a while in a company collecting survey information (on paper). Those surveys often had an identification number on the paper, which some people took as a clue to that they weren’t anonymous (as claimed). The thing is that the numbers only identified a piece of paper (so if some data looked strange one could find the paper and see if the automatic data entry had failed) and not the person answering the questions.

    So numbers, barcodes and similar on the questionnaire is not necessarily a sign that it’s not anonymous.

  16. Paul says:

    > > I then got a letter telling me I did not complete the survey and needed to.

    > Maybe *nobody* completed the survey so they sent everyone the nag letter.

    > Or maybe everybody who did complete the survey volunteered indentifying information

    It was from the local council, going to the Office of the Deputy Prime Minister, probably for purpose of quality assurance or something.

    The only identifying information was the entire A4 letter on the other side of the survey!!!

  17. Anonymous except for my IP address says:

    My employer conducted its "anonymous" employee satisfaction survey online last year.  For the first attempt, the survey was conducted via the internal Sharepoint site to which you had to log in using your company network credentials.  For some reason, the reply response was rather low and included several lucky respondents who managed to figure out that you could log in as one of the conference rooms.  The employer tried again via SurveyMonkey and got a 90+% response.  When they shared the results with the employees, the responses seemed more honest than they would have been using Sharepoint.

Comments are closed.