Tips for Conducting On-Campus Surveys

Learn about the rights of research subjects and follow procedures for submitting your protocol to the W&L Institutional Review Board.

Information to include in your initial contact e-mail or letter to participants:

  • the purpose of the survey, and how the results will be used
  • who you are representing: an on-campus organization or office, a publication, or as part of a class project
  • whether or not survey responses will be confidential (in most cases they will be; be aware that not having an expectation of confidentiality may result in biased responses or make people not want to respond at all)
  • your contact information, in case respondents have any questions or concerns

When sending survey e-mails, you should protect the privacy of survey participants by placing their e-mail addresses in the "BC:" (blind copy) line at the top of the e-mail. Put your own e-mail in the "To:" line, so you that you will receive confirmation that your e-mail to survey participants went through.

Sending out a polite follow-up reminder after a week or two to people who didn't respond to your original contact can be helpful to boost your response rate. However, it is our experience that doing more than one follow-up doesn't increase response rate appreciably, and may be irritating to the recipients.

Don't expect a 100% response rate on your survey. A typical survey response rate is 35-65%.

Think about how you are going to maintain your survey data in a database and how you want to analyze it *before* you send it out. It is a common mistake for beginning survey designers to collect much more data than they are prepared to analyze or interpret. A twenty-question survey that has a five-choice scale (Poor-Fair-Good-Excellent-No opinion) generates 100 different frequency counts! If you don't need or aren't prepared to sift through that much data, consider asking fewer questions, or giving fewer options on individual questions. Also, think carefully about asking open-ended questions; while they can be valuable sources of information, the results are often difficult to analyze.