Good survey techniques

Shawnee Love   •  
June 29, 2012

Last week I pointed out the many reasons not to do a survey of your employees.  If you still want to push ahead, here are some techniques for good surveys.

Plan ahead

Figure out what exactly you want to accomplish from your survey.  e.g., do you want to get the word out or find out how people feel about something?  Do you want to track information over time? Once you know what you want to achieve, it is easier to build a survey out from there.

Include Demographics For Metrics Not Identification

A key fear of employees doing surveys is that their employer will know who wrote the answers. That fear causes people to lie or at the very least sugarcoat their answers which totally defeats the purpose of the survey. It is natural to want to cut the data you gather, but be careful to only collect demographics on meaningful ways you will want to cut it and be sure you aren’t splitting hairs. E.g., if you have employees specify their departments but some departments have only 1 employee, you will easily be able identify who answered the questions. If you can identify who said what, you are too granular in your demographics.

Stick to Need to Know

Too many surveyors think if they are going to do a survey they should ask everything they could ever possibly want to know. Big mistake. The reason we preplan is to decide what we want to learn so that we can keep our surveys tight. People are getting surveyed out, so it is wise to focus your survey only on what you need to get out of it. Resist the temptation to ask anything else.

Design Questions Carefully

I always get irritated with leading questions, e.g., “Why is vision care important to you?”.    Vision care may not be important but now respondents are forced to answer accordingly. It is even worse if the company provides a selection of answers to choose from but neglects to offer the option “It isn’t important”.  If we continue with the vision care example, the question gathers much better information if asked as follows: “Which benefits are important to you?”

Another poor question I often see is “Do you like your extended health and dental care?” followed by the option to answer “yes”, “no” or “indifferent”.  However, combining two separate topics into the same question is not a good idea.  If the respondent likes one but not the other, it is impossible to answer truthfully.

Keep the questions simple. Double negatives and run on sentences with lots of commas are not good material for surveys. Break up long questions into bite sized pieces.

Stick to facts or opinions of the respondent and stay away from asking the employee for things the employee isn’t likely to know.  E.g., “How happy are your colleagues with our gym facilities?” is a terrible question and simply a waste of time.

Mix it up

To keep people interested and to prevent autopilot responses, it is helpful to mix up your question styles. For example, use questions requiring yes or no answers, rating answers, narrative responses, pick list questions, prioritizing questions, etc.

I often get asked if it is appropriate to allow people to skip questions. I defer back what you are trying to accomplish and let that decide whether skipping is appropriate; however, it is also important to remember that you don’t want to irritate your respondents.

Speaking of irritating respondents, I get more irritated the longer a survey is. Unless it is required by law (i.e., census) or there is really something in it for me (no, a 1 in 1000 chance of winning a $250 gift certificate doesn’t count), I have about 5 or maybe 10 minutes before I am frustrated with the survey.  Keep the questions concise. Conduct the survey online to reduce paper and handwriting requirements (and to make compiling easier). Offer answer options if you know all the available answers. Provide rating scales so respondents can simply select the level.

Communicate

Nothing will ensure your survey falls on its face faster than poor communication about why it is being done and how it will be administered and used.  Communicate before and after the survey and take the opportunity within the survey as well to ensure your employees’ questions will be answered. Topics you might want to address include:

  • What it is trying to achieve,
  • How it was developed,
  • Who or how it will be administered,
  • How long the survey will take,
  • How confidentiality and privacy will be protected,
  • What is in it for the employee (This is probably the most important of all)
  • When you will report on results/findings
  • Who to go to for questions or concerns, and
  • Thanks for participating.

Speaking of Thanks, thanks for reading my blog.  Please comment if you have anything to add to this list.