Tips on Creating Surveys to Measure Employee Satisfaction

 Employee evaluation (otherwise known as "Employee Satisfaction") on both the organizational and individual level is often one of the most forgotten about and underutilized tools in the workplace.  The sense I get is that it is overlooked because people are afraid of surveys either because they don’t know how to use them properly or they don’t know what to do with the information once they get it.  I suppose sometimes it's also true that people simply prefer "not to know" because there is a perception when it comes to feedback that less is more-which typically means that they think life will be easier if the lid can be kept on any potential dissatisfaction.  That’s a dangerous strategy though, because if it backfires, it backfires big and it can be too late to repond effectively.

Whatever the reason, surveys and evaluation don’t have to be long, cumbersome and complicated exercises that render meaningless results.  Rather, it’s just a matter of obtaining some insight that can guide you through the process as you craft a system that works best for you.

Making some decisions up front will help.  To get started, consider how you want to structure your questions.   You have some options.

   1. “Forced Choice,” is the term used when the structure of the survey “forces” respondents’ answers by steering them onto a continuum that typically spreads from “agreement” on one side to “disagreement” on the other. They are called Likert Scales and can be constructed with either an even or odd number of points.  The difference being that if you use even numbers there is essentially no middle and it “forces” people to fall on one side of a continuum representing good or bad, positive or negative. On the other hand, if you use a 3, 5, 7 or 9-point scale you make a place in the center for people to remain neutral.

   2. “Multiple Choice,” lets the respondents choose from a series of different answers.

   3. “Binary Systems,” put forth a statement that you want to have either confirmed or denied as in yes/no and true/false.  You could also have yes/no/maybe and true/false/not sure.  I suppose we could call that “Tertiary,” if someone hasn't already.  I, myself don't know.

   4. Qualitative data doesn’t allow you to quantify the answers, but it does give you a much richer sense of the information collected because it asks respondents “to explain” their answers in their own words.  The key here is to supplement the numerical data by looking for patterns and pulling them out.

Some other tips…

   1. Don’t ever create a question that combines two parts with an “and” or an “or,” because essentially what you are doing is giving people two questions at once with only an opportunity to provide one answer.  This is one of the easiest ways to skew results and muddy the outcome.

   2. Balance the quantitative with qualitative.  This helps reinforce that the numbers reflect the intentions of the respondent and also enables the compiler to crosscheck the responses if necessary. 

   3. Don’t pose all the questions exclusively with a positive or negative spin. This desensitizes respondents to the questions and you therefore run the risk of having them run down one column and checking off the same answers.  Mix it up so that they have to both agree and disagree to generate the information you are looking for.

   4. Test your questions out by asking people to respond and see if the answers give you the information you needed, and were looking for.