Should I track survey respondents? Part 2

By Marcie Levine

One of the discussions that I have with many clients is whether or not to use usernames and passwords to track employee and customer survey respondents.   On the last blog I talked about the overview thinking when deploying an internal or external survey.  This time, I’d like to discuss a few of the most common online tracking mechanisms commonly used in survey deployment?

What type of tracking is available? There is range of choices when discussing security/id options for a web-based survey:

  • You can choose to not employ any security or ID’s – you just inform your survey respondents of the URL where they will find the survey, and they complete the survey.  The pro here is that it is easy to implement, and survey respondents will know that their responses are anonymous. The cons are that a respondent will be able to take the survey more than once, and you will not be able to track who has taken the survey and who hasn’t.  Also, if you are going to send reminders to survey invitees, you will have to send reminders to everyone since you won’t know who has or has not taken the survey.
  • The second choice is to employ a ‘global password’.  In this case you would assign a password that each survey respondent will need to enter to gain access to the survey instrument.  This password would be the same for all respondents.  You might want to employ this type of security if you are concerned about people outside the company looking at your survey.  However, we believe that people outside of the company usually cannot obtain access to the surveys you will post.  Most respondents do not think that you can identify them individually using this system.  If this is the only security you use, this is a correct assumption.  The pros and cons are similar to not employing any security (as listed above).
  • The third choice for security is to use an ID login (e.g., username and/or password).  In this case you would provide this individual information to the survey respondents.  They would use this login information to gain access to completing the survey.  One reason to use this type of security is so that each respondent will be able to complete the survey only once.  The downside to this process is that the survey respondent will need to type in their login (could be considered a hassle), and they will have to type it in correctly or an error message will keep them from gaining access.  You are able to set up this system to track individual identification or not.  If you do use this system, most survey respondents will believe you are tracking who they are (whether you are or not), so be sure to let respondents know how you plan to use the data.
  • The fourth choice for security is to attach an ID to the end of the URL location of the survey.  This ID can be an anonymous ID (not attached to any demographic information), or attached to information about the respondent.  When the respondent clicks on the URL, the ID can be automatically placed in the first answer field.  After the data is collected, the demographic information about the respondent can be merged with the survey answer files.  This security is an easy way for you to collect identifying information about your survey respondent without asking them to enter any additional information.  Many respondents know that you are identifying them in this fashion; some even take off the ID from the end of the URL.  They can still take the survey, but their ID will not be entered in the first field of the response data set.

The conversation about whether to track respondents includes trying to understand how much potential bias you are willing to introduce to your survey results balanced with collecting enough data to make your results relevant.  The less the security, the more chance there is for respondents to incorrectly self-identify any demographics, and the more able they will be to take the survey more than once. The more security you put in place, the more you will be able to track the individual respondents, and the less anonymous your responses will be.

We hope that this article promotes some critical thinking about this issue.  Feel free to contact SurveyConnect if you’d like to discuss a specific issue you have with web-based survey security.

Ready to get started?

About the author

Marcie Levine

Marcie Levine

Prior to founding SurveyConnect, Marcie had more than twelve-years of experience in HR, both as a consultant and in several corporate positions. As an HR consultant, she worked on a variety of client assignments, including the design and implementation of surveys. She created SurveyConnect to streamline and simplify the process of survey creation and administration. Marcie has published articles about 360 feedback process in trade magazines, including the International Society for Performance Improvement, Selling Power, T+D Magazine, and Quirk’s Marketing Research Review.